Business and other application programs execute in a wide variety of operating environments. Before deploying an application program into the operating environment in which the application programs is intended to execute, it is often desirable to test the application program to determine its suitability for the operating environment and to discover any problems associated with the application program before any undesirable physical, economic, or other damage may occur. For example, it might be desirable to test an application program that acts upon or using dates expressed in a format having a two-digit calendar year, and for which conversion to a format having a four-digit year is performed, due to potential ambiguities associated with dates on and after Jan. 1, 2000 and potential problems that may arise because of these ambiguities.
As the number and complexity of application programs and operating environments continue to increase, along with the potential damage that defects and problems associated with application programs may cause in some circumstances, techniques for testing application programs have become increasingly important. A known technique for testing an application program includes generating a simulation of the operating environment in which the application program is intended to execute, executing the application program within the simulated operating environment, recording the testing results, and discarding the simulated operating environment. If the application program is to be tested within a different operating environment, a simulation of this different operating environment must be generated and the application program retested. Although such techniques may be acceptable in some circumstances, they are often inadequate in cases in which a large number of application programs are to be tested in a large number of operating environments, due to the time and costs associated with generating the simulated operating environments. Moreover, multiple users are often unable to concurrently access a particular simulated operating environment, resulting in additional delays and costs. These and other disadvantages make previous techniques for testing application programs inadequate for many purposes.